As a method to visualize a fetal image, ultrasonography devices are mounted with a three-dimensional (3D) display function using volume rendering for example. Upon 3D visualization, setting a 3D region of interest (hereinafter referred to as 3D-ROI) for limiting a range where rendering processing is applied is required. If this 3D-ROI setting is not performed as accurately as possible, there is a problem that the face of a fetus cannot be preferably displayed in an image obtained in 3D attributable to floating substances in the amniotic fluid or the placenta. Therefore, accurate setting can be performed with subdivided setting items for the 3D-ROI. As a result, however, operations for the setting are disadvantageously extremely complicated.
In recent years, technique to automatically set or correct this 3D-ROI is disclosed. PTL 1 discloses technique to detect a contour of a fetus from contour points that are points on the contour of the fetus specified by a doctor and to set a 3D-ROI. PTL 2 discloses technique to implement a smooth convex surface or a concave surface by generating a starting plane of rendering processing by a spline function since a shape of the amniotic fluid region that is a region between a fetus and the placenta is empirically substantially a convex surface or a concave surface and to obtain appropriate 3D display.